General neovim hack for developers

[A7Lavinraj]

Contributing guidelines

• I have read CODE_OF_CONDUCT.md

Module(s)

none of the above

Question

Hi there! This discussion is not related with any of the mini modules but for a little advice request. You know during plugin development there are several scenarios where we just add print statements to debug things out at a particular stage of plugin flow.

Sometimes messages are too long to fit in the screen height. So what did you do to handle that situation to move through those long messages, because you cannot do much to move around other than u,d,j,k to move around.

In some of the streams of tjdevries I found that he can open messages in a split with a Messages command which looks like custom made. I tried to ask same question on his neovim config but haven't got any response.

I though maybe you might also face this issue. If you have any advice to handle that please let me know or just close this discussion

BTW I recently start using mini.pick with mini.extras they are amazing thank you!

[echasnovski]

Hi! Thanks for kind words!

Up until Neovim 0.12 it was indeed mostly :messages and navigate through it. The u,d,j,k (and G + g) was enough for me. The thing is, checking out messages is not that frequent operation for me.

You know during plugin development there are several scenarios where we just add print statements to debug things out at a particular stage of plugin flow.

Printing messages is not the best approach of doing this. A better way is to add some global _G.log = {} array and append to it the data you want to see during debugging (don't forget to vim.deepcopy if there are mutating tables).

As this is a frequent operation for me, I had custom helpers which not so long ago I added to 'mini.misc'. Those are various log_xxx() functions. Together with la prefixed snippet and dedicated mappings, this feels pretty smooth right now.

To answer the Messages command question, it should be a more or less concise implementation. See :h nvim_exec2() to capture the output of any command (like messages). Later post processing and maybe opening in a separate buffer is left as a homework :)